ML Infrastructure Engineer

What you'll do

• Profile and analyze GPU performance at system and kernel level; compare platforms, architectures, and software stacks.
• Debug and optimize ML workloads, test new GPU clusters, and run experiments on system configurations and interconnects.
• Build tools and dashboards to visualize performance metrics, bottlenecks, and trends; contribute to internal tooling and best practices.

What they ask for

• Strong theoretical understanding of machine learning and performance characteristics of large neural-network training and inference.
• Experience with modern deep-learning frameworks, including PyTorch, JAX, Megatron-LM, and Tensort-LLM.
• Understanding of the GPU stack, including CUDA, NCCL, drivers, and relevant libraries.
• Familiarity with containerized environments such as Docker and Kubernetes; strong communication and ability to work independently.

Nice to have

• Familiarity with LLM inference frameworks such as vLLM, SGLang, and TensorRT; experience with Python and performance profiling tools.
• Familiarity with AWS, GCP, or Azure ML; contributions to open-source ML benchmarking tools.

Tools

CUDA, ROCm, PyTorch, JAX, Megatron-LM, Tensort-LLM, NCCL, Docker

AI focus

Benchmarks and optimizes GPU hardware for machine-learning and AI workloads, including deep-learning training and inference.